n=1

We are constantly talking about all of these people in this country who are dying from medical errors. According to a recent publication, it is the third leading cause of death in the US. But who is doing all this killing, because it's certainly not me, right? I haven't killed anyone because of any mistakes. But statistically, I have to be a part of this number. I made some little something somewhere, that when added up on a balance sheet, pushed someone from the "alive" column" into the "dead" column. But it's hard to know what to do about this, personally. As a system, there are lots of things that organizations do, like tracking and labeling, human factors engineering, checklists, etc. etc. But what can I do?

In medicine, we talk a lot about "n=1" studies. In a research protocol, n is the number of participants in a study. The bigger the n, the more powerful the study to make a conclusion. But in practice, it is those n=1 studies that influence our behavior. I missed a cholangiocarcinoma whose only significant finding was a mildly elevated alkaline phosphatase. That was my only clue, and I dithered. And now, I'm suspicious of every alk phos elevation, constantly wary of missing this diagnosis. But it is a fairly uncommon cancer. I'll see tons of breast and colon and prostate and lung cancers in order to see one more cholangiocarcinoma. Of the 454 cancers found yearly per 100,000 population, cholangiocarcinoma makes up 1-2 total. There will be 452+ other cancers that I am going to see before I see another one. But here I am, fretting about a mildly elevated alk phos, wondering if I should get an MRCP.

A friend of mine described being a doctor as being continuously haunted by ghosts. There is always some ghost of a previous patient hanging over you, reminding you of your mistakes and failings. And it takes a fair bit of bravery to count up those ghosts and look for any patterns.